Service Level Agreement
The strict mathematical guarantees defining our uptime parameters and automated credit distribution.
Threshold: ≤ 43.8 minutes of downtime/month
Effective Date: March 2026 • Last Updated: May 20, 2026
TL;DR: The Core Parameters
Live Global Pulse- 99.9% Metric Target: We guarantee the continuous routing of packets equivalent to 99.9% availability per operational month, directly tracked via status.gozenhost.com.
- Automated Reconciliation: Valid outages trigger direct billing credits up to 100% of your specific monthly base instance cost.
- Downtime Definition: "Downtime" means 5+ consecutive minutes of zero packet delivery to your instance, verified via status.gozenhost.com. Scheduled maintenance announced in advance does not count.
- 15-Minute Emergency Response: For production-down incidents on paid plans, an engineer acknowledges and triages your ticket within 15 minutes of submission, 24/7. In practice we usually respond faster. 15 minutes is the worst-case commitment, not the target.
- Non-Covered Edges: Customer-initiated reboots, OS misconfigurations, DDoS attacks targeting your specific IP layer, and announced maintenance windows are excluded.
1. Availability Monitoring Protocol
We actively monitor the global pulse of our datacenters across independent geographic nodes. The mathematical "Uptime Percentage" is calculated by taking the total minutes in a given billing month minus the aggregate Minutes of Downtime, divided by the total minutes in the month, multiplied by 100.
"Downtime" is precisely defined as a sustained period where GOZEN HOST boundary/edge routers register zero packet delivery to an operational virtual instance node for more than five (5) consecutive minutes. Verification of the pulse state is globally accessible via status.gozenhost.com.
2. Credit Disbursement Matrix
In the event we fail to satisfy the 99.9% infrastructure agreement, your account accrues SLA credits structured on the operational deficit. Credits represent an invoice discount for the following billing cycle, not a cash refund.
| Logged Monthly Uptime (%) | Equivalent Max Downtime | Credit (% of Month Fee) |
|---|---|---|
| ≥ 99.9% | ≤ 43.8 minutes | 0% (SLA Met) |
| 99.5% to 99.89% | up to 3.6 hours | 10% Credit |
| 99.0% to 99.49% | up to 7.2 hours | 25% Credit |
| 95.0% to 98.99% | up to 36.0 hours | 50% Credit |
| < 95.0% | > 36.0 hours | 100% Credit |
3. Credit Claim Procedure
To request an SLA credit, follow these steps:
- Open a Support Ticket within 14 days of the disrupted billing cycle. Select the category: Billing → SLA Credit Claim.
- Include required details: the affected service/hostname, start and end timestamps of the outage, and a reference link to the status.gozenhost.com incident (if applicable).
- Verification: Our operations team cross-references your claim against internal telemetry and status page records. Global routing failures that trigger mass-automated credits do not require a manual claim.
- Credit applied: Verified credits are applied to your next billing cycle within 5 business days. SLA credits discount future invoices only. They cannot be converted to cash refunds.
4. Architectural Exclusions
The SLA calculation strictly excludes conditions categorized as exogenous or user-induced. These include:
- Planned Maintenance (Status Portal): Scheduled maintenance windows affirmatively announced on status.gozenhost.com prior to execution are intentionally excluded from "Downtime" tracking.
- Environment Corruption: Container downtime caused by recursive loops, OOM (Out Of Memory) process kills instantiated by your application, failed kernel updates, or custom IP table blockades you configured.
- Direct External Attacks: Volumetric DDoS floods originating from external networks specifically targeting your un-proxied primary or secondary IP assignments.
- Force Majeure / Carrier Faults: Natural disasters severing transatlantic cabling or regional ISP outages preventing downstream connections from reaching our edge routers.
5. Incident Response Commitment
In addition to the availability guarantee, we commit to a 15-minute response window for production-impacting incidents reported by customers on any paid plan, 24 hours a day, 7 days a week. An on-call engineer acknowledges the ticket, confirms receipt, and assigns an initial triage status within that window. In practice the median response in our ticket logs sits well below 15 minutes. The commitment defines the maximum acceptable wait, not the target.
What "response" means here: an acknowledgement and triage assignment from a human engineer, not a complete resolution. Time-to-resolution varies with the nature of the incident (kernel panic, network anomaly, application bug, dependency outage) and is not bounded by this clause.
What qualifies as a production-impacting incident: your hosted services are unreachable, returning systemic errors, or measurably degraded in a way that affects end users. Examples include unreachable IPs, sustained 5xx responses, database connection refusals, mail queue stalls, and broken SSL renewals on live domains.
What does not qualify: general configuration questions, billing inquiries, feature requests, account administration, and pre-sales conversations. Those follow our standard support response targets and are not subject to the 15-minute window.
How to invoke it: open a support ticket and select the category Technical → Production Outage. The 15-minute clock starts at ticket submission timestamp. Live chat is monitored during business hours as a secondary channel; for after-hours emergencies the ticket path is the authoritative one because it generates pager events for the on-call rotation.
If we miss the 15-minute response on a qualifying production-down incident, you may include that fact in any subsequent SLA credit claim filed under Section 3. Repeated response-time failures are treated as a separate operational defect and reviewed by the head of operations.
Related Legal Documents
Need to File an SLA Claim?
If you believe your deployment suffered an outage beyond the parameters outlined, our operations team will review your server logs and issue credits immediately upon verification.
